Academic planning
Problem: Dates, courses, evaluations and study time were scattered.
Solution: Academic calendars, weekly planners, prioritization by urgency or difficulty and suggested study blocks.
Back to home
ExploraTec Showcase
In the final exam for XTECC21 Create your AI Agent, students in ExploraTec UDD built small, working demos from problems they knew firsthand: study planning, rubrics, budgets, habits and workflows from their own fields.
EspañolCourse data
The course brought together 30 students, 22 projects and 14 majors. Most students were early in their university path, and they arrived with different levels of technical experience. That is what made the work interesting: they used Codex to turn judgment from their own field into a prototype they could test.
Majors represented
Commercial Engineering, Global Business Administration, Architecture, Dentistry, Medical Technology, Computer Science and AI, Psychology, Computer Science and Technological Innovation, Law, Civil Engineering, Political Science and Public Policy, Industrial Engineering, Kinesiology, Design.
Challenges students found
The projects show the same practical pattern: students had to turn vague friction into a flow that could be tested.
Instructions and material
Problem: Rubrics, assignments, notes and summaries were long or unclear.
Solution: Apps that simplify instructions, extract requirements and generate study guides, flashcards or review plans.
Wellbeing and habits
Problem: Procrastination, overload, anxiety, lack of sleep and daily time loss.
Solution: Focus sessions, preventive alerts, realistic schedules, outfit automation and personal organization.
Finance and consumption
Problem: Students struggled to control expenses, plan purchases or decide with limited budget.
Solution: Financial dashboards, expense classifiers, optimized shopping lists and price-benefit recommendations.
Specific career tools
Problem: Health, design, advertising and academic evaluation had discipline-specific needs.
Solution: Generated design briefs, health study simulators, campaign analysis and rubric assistants.
What Codex made possible
Codex worked best as a practical loop: describe the problem, build a version, test it, then make the next instruction sharper.
A faster path from idea to demo
Students described a problem in plain language, then used Codex to turn it into screens, data, rules and a working first version.
Iteration became part of the work
They tested what broke, narrowed the scope and came back with better instructions. The prototype improved because the student kept making choices.
Software could be personal
Several projects were not business ideas. They were small tools built for one person's routine, calendar, budget, closet or class. Codex made that kind of custom software realistic.
The student still owned the judgment
Codex could produce code, but the student chose the problem, set the limits and decided whether the result made sense.
Student projects
Calendaris
Nicolás Soto
Computer Science and Artificial Intelligence
A university calendar for courses, evaluations, events and suggested study blocks.
StudyBrain
Nicolás Campos
Computer Science and Technological Innovation
An academic platform for grades, topics, evaluations and weekly study plans.
BriefLab
Amelia Troncoso
Graphic Design
A fictional brief generator for design practice and portfolio building without real clients.
FinEx
Sebastián Ramírez
Industrial Engineering
A local-first personal finance app that imports Gmail notices, classifies expenses and visualizes budgets.
Fit Me
Diego Rivera y Valentina Castillo
Psychology and Dentistry
A Codex automation that prepares seven weekly outfits from real clothing, style references and weather.
RubricaCheck
José Tomás Cifuentes
Civil Engineering
A web app that turns rubrics and assignment instructions into a clearer structure for submissions.
For more information, contact the professor: @aastroza